The overarching purpose of this proposal is to redress limitations in current research on HIV sexual risk. Previous research focuses on deliberative, belief-based attitudes toward sexual risk behaviors. Sexual contexts, however, epitomize situations in which systematic retrieval of one's attitudes toward condoms and HIV prevention might often be minimal. Recent research (Marsh et al., 2001, 2005;Czopp et al., 2004) suggests that in some sexual contexts behavior is better predicted by implicit attitudes. Implicit sexual attitudes are evaluative responses that are automatically and effortlessly evoked by cues in a sexual situation and involve feelings rather than verbally articulated thoughts. Response latency methods, specifically computerized priming-type and implicit association tasks, are used to assess implicit condom attitudes. Implicit measures have great promise and have yielded intriguing, though inconsistent, results in the area of HIV risk. These inconsistencies may be a result of limitations in assessing behavior. To address these limitations, in the current research, we use cutting-edge methods of immersive virtual reality technology (VR) to test key theoretical predictions about the relation between implicit condom attitudes and impulsive HIV risk-related sexual behavior. In doing so, we propose to resolve theoretical and methodological problems with measuring implicit attitudes in artificial contexts and resolve previous contradictory findings resulting from "unembodied" self- report behavior measures. One qualitative study serves to first validate behavioral responses in VR and examine the dyamics of VR processes in-depth, comparing behavioral and psychological responses elicited by VR, role-playing, and written scenarios. The study will be conducted at field sites and in university laboratories with participants that include two community samples, and will include biological concomitants of risk exposure included for validation. Four experiments conducted at an urban, community health service site and in university laboratories are designed to test the personal, social, and environmental factors that cue impulsive behavior. In each experiment, implicit measures of attitudes are hypothesized to predict participants'HIV risk- related responses in impulsive situations over and above explicit measures. In contrast, explicit, intention- based verbalizable attitudes predominantly should explain behavior in low impulsivity situations. One of the social factors experiments examines social cues involving cultural risk factors unique to Latina women in an urban population (partner motives);similar cues are examined in a non-Latina sample. Two additional laboratory and field experiments will test the hypothesis that changing implicit attitudes will change impulsive responses in virtual reality, and validate that responses in VR predict subsequent risk-related behavior outside of VR.